Unmanaged datasets can come in various structures and forms. For example, an acquisition system with little or no documentation may be the source of unmanaged datasets. Yet another source of unmanaged data sets is the rise of Big Data where data from a variety of sources is loaded for example into a Hadoop platform. In such an environment, for example, text analytics could find example in free text interesting values which can be annotated with metadata producing value sets with a bit of metadata associated with it. Understanding the datasets in such systems may be the first step to determine if there is a potential of data consolidation reducing IT costs or to understand if the data assets are properly managed.